{smcl}
{com}{sf}{ul off}{txt}{.-}
      name:  {res}<unnamed>
       {txt}log:  {res}C:\Users\nkalmoe\Dropbox\Research\A Publishing\Finished\PSRM - VCM Partic\Kalmoe - Fig 2, Study 2.smcl
  {txt}log type:  {res}smcl
 {txt}opened on:  {res}14 Jun 2017, 10:53:17

{com}. do "C:\Users\nkalmoe\Dropbox\Research\A Publishing\Finished\PSRM - VCM Partic\PSRM - Kalmoe Fig 2, Study 2.do"
{txt}
{com}. //PSRM - Kalmoe Figure 2, Study 2
. 
. ***Remember the order of variables is X, Z, W. ***
. ***On the graph, W represents the different lines.  ***
. 
. ***Nathan Note: This file attempts to combine parts of the 2- and 3-way ME plots from Golder. 
. ***The goal is to show the effect of violent metaphors on the change in partisan polarization on issues.
. ***Estimates 3-way model, subtracts difference between 2 lines, calculates SEs similar to 2-way model
. 
. * y = x z w xz xw zw xzw*
. * x(Treat), z(aggress), w(PID), y(dependent variable)
. 
. reg particnovote viol aggress commitment violxaggress violxcommitment aggressxcommitment violxaggxcommit

      {txt}Source {c |}       SS       df       MS              Number of obs ={res}     396
{txt}{hline 13}{char +}{hline 30}           F(  7,   388) ={res}    6.21
    {txt}   Model {char |} {res} 2.17838271     7  .311197531           {txt}Prob > F      = {res} 0.0000
    {txt}Residual {char |} {res} 19.4477283   388  .050123011           {txt}R-squared     = {res} 0.1007
{txt}{hline 13}{char +}{hline 30}           Adj R-squared = {res} 0.0845
    {txt}   Total {char |} {res}  21.626111   395  .054749648           {txt}Root MSE      = {res} .22388

{txt}{hline 20}{c TT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{col 1}       particnovote{col 21}{c |}      Coef.{col 33}   Std. Err.{col 45}      t{col 53}   P>|t|{col 61}     [95% Con{col 74}f. Interval]
{hline 20}{c +}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{space 15}viol {c |}{col 21}{res}{space 2} .1939459{col 33}{space 2} .1119052{col 44}{space 1}    1.73{col 53}{space 3}0.084{col 61}{space 4}-.0260706{col 74}{space 3} .4139623
{txt}{space 12}aggress {c |}{col 21}{res}{space 2} .2756975{col 33}{space 2} .2326499{col 44}{space 1}    1.19{col 53}{space 3}0.237{col 61}{space 4}-.1817147{col 74}{space 3} .7331098
{txt}{space 9}commitment {c |}{col 21}{res}{space 2} .4895185{col 33}{space 2} .1924075{col 44}{space 1}    2.54{col 53}{space 3}0.011{col 61}{space 4} .1112267{col 74}{space 3} .8678104
{txt}{space 7}violxaggress {c |}{col 21}{res}{space 2}-.4802596{col 33}{space 2} .2762316{col 44}{space 1}   -1.74{col 53}{space 3}0.083{col 61}{space 4}-1.023358{col 74}{space 3} .0628384
{txt}{space 4}violxcommitment {c |}{col 21}{res}{space 2}-.2396179{col 33}{space 2} .2229588{col 44}{space 1}   -1.07{col 53}{space 3}0.283{col 61}{space 4}-.6779765{col 74}{space 3} .1987407
{txt}{space 1}aggressxcommitment {c |}{col 21}{res}{space 2}-.5539822{col 33}{space 2} .5066403{col 44}{space 1}   -1.09{col 53}{space 3}0.275{col 61}{space 4}-1.550086{col 74}{space 3} .4421217
{txt}violxaggxcommitment {c |}{col 21}{res}{space 2} .9043142{col 33}{space 2} .5998886{col 44}{space 1}    1.51{col 53}{space 3}0.133{col 61}{space 4}-.2751249{col 74}{space 3} 2.083753
{txt}{space 14}_cons {c |}{col 21}{res}{space 2} .0591482{col 33}{space 2} .0952407{col 44}{space 1}    0.62{col 53}{space 3}0.535{col 61}{space 4}-.1281043{col 74}{space 3} .2464007
{txt}{hline 20}{c BT}{hline 11}{hline 11}{hline 9}{hline 8}{hline 13}{hline 12}
{res}{txt}
{com}. 
. #delimit ;
{txt}delimiter now ;
{com}. *Estimating marginal effect of treatment, conditional on TA and PID*;
. egen zmin = min(aggress);
{txt}
{com}. egen zmax = max(aggress);
{txt}
{com}. gen z0 = (((_n-1)/(20-1))*(zmax-zmin))+zmin in 1/20;
{txt}(392 missing values generated)

{com}. generate Z=z0;
{txt}(392 missing values generated)

{com}. *     ****************************************************************  *;
. *       Generate the values of W for which you want to calculate the    *;
. *       marginal effect (and standard errors) of X on Y.                *;
. *     ****************************************************************  *;
. scalar W0=0;
{txt}
{com}. scalar W1=1;
{txt}
{com}. *     ****************************************************************  *;
. *       Grab elements of the coefficient and variance-covariance matrix *;
. *       that are required to calculate the marginal effect and standard *;
. *       errors.                                                         *;
. *     ****************************************************************  *;
. matrix b=e(b);
{txt}
{com}. matrix V=e(V);
{txt}
{com}. scalar b1=b[1,1];
{txt}
{com}. scalar b2=b[1,2];
{txt}
{com}. scalar b3=b[1,3];
{txt}
{com}. scalar b4=b[1,4];
{txt}
{com}. scalar b5=b[1,5];
{txt}
{com}. scalar b6=b[1,6];
{txt}
{com}. scalar b7=b[1,7];
{txt}
{com}. scalar varb1=V[1,1];
{txt}
{com}. scalar varb2=V[2,2];
{txt}
{com}. scalar varb3=V[3,3];
{txt}
{com}. scalar varb4=V[4,4];
{txt}
{com}. scalar varb5=V[5,5];
{txt}
{com}. scalar varb6=V[6,6];
{txt}
{com}. scalar varb7=V[7,7];
{txt}
{com}. scalar covb1b4=V[1,4];
{txt}
{com}. scalar covb1b5=V[1,5];
{txt}
{com}. scalar covb1b7=V[1,7];
{txt}
{com}. scalar covb4b5=V[4,5];
{txt}
{com}. scalar covb4b7=V[4,7];
{txt}
{com}. scalar covb5b7=V[5,7];
{txt}
{com}. scalar list b1 b2 b3 b4 b5 b6 b7 varb1 varb2 varb3 varb4 varb5 varb6 varb7 
>            covb1b4 covb1b5 covb1b7 covb4b5 covb4b7 covb5b7;
{txt}        b1 = {res} .19394586
{txt}        b2 = {res} .27569755
{txt}        b3 = {res} .48951853
{txt}        b4 = {res}-.48025957
{txt}        b5 = {res}-.23961786
{txt}        b6 = {res}-.55398224
{txt}        b7 = {res} .90431419
{txt}     varb1 = {res} .01252278
{txt}     varb2 = {res} .05412598
{txt}     varb3 = {res} .03702065
{txt}     varb4 = {res} .07630388
{txt}     varb5 = {res} .04971062
{txt}     varb6 = {res} .25668442
{txt}     varb7 = {res} .35986634
{txt}   covb1b4 = {res}-.02475607
{txt}   covb1b5 = {res}-.02286712
{txt}   covb1b7 = {res} .04562007
{txt}   covb4b5 = {res} .04612414
{txt}   covb4b7 = {res}-.14718724
{txt}   covb5b7 = {res}-.10464422
{txt}
{com}.            *     ****************************************************************  *;
. *       We want to calculate the marginal effect of X on Y for all      *;
. *       Z values of the modifying variable Z. We also want to          *;
. *       calculate this marginal effect as Z changes for specific values *;
. *       of the second modifying variable W.  In the code below, we      *;
. *       calculate the marginal effect of X on Y for all values of Z     *;
. *       when W=0, when W=1, when W=2, and when W=3.                     *;
. *     ****************************************************************  *;
. gen conb0=b1+b4*Z+b5*W0+b7*(Z*W0);
{txt}(392 missing values generated)

{com}. gen conb1=b1+b4*Z+b5*W1+b7*(Z*W1);
{txt}(392 missing values generated)

{com}. gen conbdif=conb1-conb0;
{txt}(392 missing values generated)

{com}. *     ****************************************************************  *;
. *       Calculate the standard errors for the marginal effect of X on Y *;
. *       for all Z values of the modifying variable Z. Do this for the  *;
. *       case when W=0, when W=1, when W=2, and when W=3.                *;
. *     ****************************************************************  *;
. gen conse0=sqrt(varb1
>                + varb4*(Z^2) + varb5*(W0^2) + varb7*(Z^2)*(W0^2)
>                + 2*Z*covb1b4 + 2*W0*covb1b5 + 2*Z*W0*covb1b7 + 2*Z*W0*covb4b5
>                + 2*W0*(Z^2)*covb4b7 + 2*(W0^2)*Z*covb5b7);
{txt}(392 missing values generated)

{com}. gen conse1=sqrt(varb1
>                + varb4*(Z^2) + varb5*(W1^2) + varb7*(Z^2)*(W1^2)
>                + 2*Z*covb1b4 + 2*W1*covb1b5 + 2*Z*W1*covb1b7 + 2*Z*W1*covb4b5
>                + 2*W1*(Z^2)*covb4b7 + 2*(W1^2)*Z*covb5b7);
{txt}(392 missing values generated)

{com}.                                                    gen consedif=sqrt(conse0^2 + conse1^2);
{txt}(392 missing values generated)

{com}.                                                                 gen a=1.645*consedif;
{txt}(392 missing values generated)

{com}.  gen upper=conbdif+a;
{txt}(392 missing values generated)

{com}.  gen lower=conbdif-a;
{txt}(392 missing values generated)

{com}.                         graph twoway line conbdif   Z, clwidth(medium) clcolor(blue) clcolor(black)
>         ||   line upper  Z, clpattern(dash) clwidth(thin) clcolor(black)
>         ||   line lower  Z, clpattern(dash) clwidth(thin) clcolor(black)
>         ||   ,   
>                  xlabel(0(.2)1, labsize(5)) 
>              ylabel(-1(.5)2, labsize(5))
>                  legend(off)
>              yline(0, lcolor(black))   
>              xtitle(Trait Aggression, size(5)  )
>              xsca(titlegap(2))
>              ysca(titlegap(2))
>              ytitle(Marginal Treatment FX on Impact of Motives, size(4))
>              scheme(s2mono) graphregion(fcolor(white));
{res}{txt}
{com}.              drop zmax zmin z0 Z conb0 conb1 conbdif conse0 conse1 consedif a upper lower;
{txt}
{com}. 
{txt}end of do-file

{com}. log close
      {txt}name:  {res}<unnamed>
       {txt}log:  {res}C:\Users\nkalmoe\Dropbox\Research\A Publishing\Finished\PSRM - VCM Partic\Kalmoe - Fig 2, Study 2.smcl
  {txt}log type:  {res}smcl
 {txt}closed on:  {res}14 Jun 2017, 10:53:26
{txt}{.-}
{smcl}
{txt}{sf}{ul off}